The Preposition 'from'

Having just now discovered I missed National Preposition Day, I offer a post relating to the preposition from and my dissertation.

It has long been noted that the English preposition from most typically occurs with Source (Huddleston and Pullum 2002; Van Valin and LaPolla 1997; Jolly 1991; Clark and Carpenter 1989a; Clark and Carpenter 1989b; Quirk 1985; Vestergaard 1977; Wood 1967).

(1) a. Chris returned from California.
b. Hide took the book from Atsuko.
c. Mike drove from Buffalo to Toronto.

There are some uses, however, where it appears to occur with Goals and Themes as well.

(2) a. The fence blocked the car from the driveway.
b. The tent shielded the kids from the rain.

When from occurs with verbs denoting barrier events like bar, ban, block, shield it can marks NPs representing either the unattained goal of the entity being blocked, or the restrained theme which failed to attain its goal. Interestingly, it can also mark VPs as in (3):

(3) The judge barred the journalists from entering the courtroom.

In my first qualifying paper (SUNY Buffalo’s linguistics department uses qualifying papers in lieu of a master’s thesis) I argued that from acts not as a preposition, but rather as a complementizer with barrier verbs.

The class of English “barrier verbs” (as originally sketched by Len Talmy) are negative causative object control verbs which encode the relationships between a goal directed participant (or “agonist”), its goal and a barrier participant (or “antagonist”). They are negative verbs in the sense of Laka 1994: they encode the negation of an event. Think of the verb neglect. If you neglect to do X, then X did not happen. With barrier verbs, if you ban X from doing Y, then Y did not happen.

These verbs fall into the following general constructional template:

NP₁ verb NP₂ from NP₃/VP.

In this construction, the subject of a barrier verb (NP₁) acts as the barrier (either directly or indirectly) to the achievement of a goal event (NP₃/VP) by a goal-directed participant NP₂).

I have recently discovered that Idan Landau has a detailed analysis of negative verbs in Hebrew which, extended to English prevent, suggests the complementizer interpretation as well. Though we have very different theoretical frameworks, I think we share some conclusions.

I’m interested in a variety of the phenomenon associated with barrier verbs (including the potential for a coercion analysis of the NP replacement of complement VPs)

. UPDATE: Barbara Partee discusses negative events in this Language Log post.